-
The most likely reason rocBLAS API limitations (some parameters are passed as ints, which means 2G limit). Another possible reason overflow due to is natural limitations of the FP16 datatype, but this…
-
Hi, I have a problem with using Vicuna13b-v1.3 to make an **inference with multi-GPU**. Could anyone please provide an example of code used for multi-GPU inference without the CLI? On the other hand, …
-
Please let me know how to reproduce the SSIM and PSNR value of NHHaze dataset given in paper. I am getting on .509 as SSIM value
-
### 🐛 Describe the bug
When calling conv_transpose2d with groups > 1, I run into the following error:
```
r = torch.nn.functional.conv_transpose2d(a, wts, stride=stride, padding=padding, dila…
-
## Habitat-Lab and Habitat-Sim versions
Habitat-Lab: vx.x.x or master?
master
Habitat-Sim: vx.x.x or master?
v2.4.0
Habitat is under active development, and we advise users to restrict themselves…
-
anyone project dev can share their cudnn version and their torch config?
```bash
python -m torch.utils.collect_env
```
it keeps crashing
```bash
\site-packages\torch\nn\modules\conv.py:306: …
-
Currently `TilizeWithValPadding` only supports width sharding, convs are often height sharded so they're unable to use `TilizeWithValPadding` without going to interleaved which is expensive.
-
Hi, In the Adaptive Attention model, your inputs to the forward function are:
`def forward(self, h_out, fake_region, conv_feat, conv_feat_embed)`
Where in the code have you extracted those features…
-
### Describe the bug
![ezgif-7-84762494ee](https://github.com/user-attachments/assets/838a1658-24d0-4d13-92b2-ff8991da945f)
### Expected behavior
_No response_
### Steps to reproduce the…
-
In networks.py, when init class RED_CNN, only 4 nn.Conv2d were instantiated.
```
def __init__(self, out_ch=96):
super(RED_CNN, self).__init__()
self.conv_first = nn.Conv2d(1, o…